Skip to main content

FrootAI — AmpliFAI your AI Ecosystem Get Started

All Solution Plays

Play 73

Waste & Recycling Optimizer

Medium Designed

AI-powered waste management — material classification, route optimization, contamination detection.

AI-powered waste management system combining computer vision material classification, collection route optimization, and contamination detection. Azure AI Vision identifies waste types (recyclable/compostable/landfill) from camera feeds, OpenAI generates route optimization considering vehicle capacity and pickup schedules, IoT Hub tracks bin fill levels, and Cosmos DB stores recycling rate analytics for municipality dashboards.

Architecture Pattern

Vision + IoT waste management: material classification → route optimization → contamination alerts

Azure Services

Azure AI VisionAzure OpenAIAzure IoT HubContainer AppsCosmos DB

DevKit (.github Agentic OS)

  • agent.md — root orchestrator with builder→reviewer→tuner handoffs
  • 3 agents — Waste Builder (gpt-4o), Reviewer (gpt-4o-mini), Tuner (gpt-4o-mini)
  • 3 skills — deploy (186 lines), evaluate (132 lines), tune (230 lines)
  • 4 prompts — /deploy, /test, /review, /evaluate with agent routing
  • .vscode/mcp.json — FrootAI MCP with Custom Vision + Maps inputs + envFile

TuneKit (AI Config)

  • config/openai.json — classification and optimization prompts
  • config/waste.json — material categories, vehicle capacity, pickup windows
  • config/guardrails.json — classification confidence thresholds
  • evaluation/eval.py — Classification accuracy >90%, Route efficiency >85%

Tuning Parameters

Material categoriesClassification confidence thresholdVehicle capacity constraintsPickup time windowsContamination sensitivity

Machine evidence

FrootAI evidence lifecycle

This is an internal evidence maturity label, not third-party certification, accreditation, legal compliance, or a production guarantee. Missing or expired evidence demotes automatically; catalog claims cannot promote a play.

Designed
designed
build verified
evaluation verified

This play currently has design evidence only. A runnable scenario, endpoint evaluation, and build receipts are the next contiguous gates.

Loading architecture and cost model…

Repo Intelligence

v1

A no-clone, revision-pinned map for agents and humans. Observed evidence is separated from inferred flow so the output stays useful without pretending to be a full call graph.

Indexing bounded repository evidence…